THE JOURNAL OF FINANCE • VOL. LX, NO. 6 • DECEMBER 2005
CEO Compensation, Change, and
Corporate Strategy
JAMES DOW and CLARA C. RAPOSO∗
ABSTRACT
CEO compensation can inf luence the kinds of strategies that firms adopt. We argue
that performance-related compensation creates an incentive to look for overly ambitious, hard to implement strategies. At a cost, shareholders can curb this tendency by
precommitting to a regime of CEO overcompensation in highly changeable environments. Alternatively shareholders can commit to low CEO pay, although this requires
a commitment mechanism (either by the board of the individual company, or by the
society as a whole) to counter the incentive to renegotiate upwards. We study the
conditions under which the different policies for CEO compensation are preferred by
shareholders.
We have created a cult of leadership that far exceeds anything that existed decades ago. . . What we are getting now, very dangerously, is what
I call a dramatic style of managing; the great merger, the great downsizing, the massive brilliant new strategy. . . So we get all these massive
mergers, fire, brimstone and drama, because you can’t say to the stock analysts, “we’re getting our logistics all straightened out, we’re going to be
much more efficient at throughput to the customer.” They start to yawn.
(Mintzberg (2000, p. 31))
The problems resulting from separation of ownership and control have long
been recognized in the corporate governance and corporate finance literature
(See, e.g., Berle and Means (1932), Jensen (1986), Hart (1995), Shleifer and
Vishny (1997)). Mintzberg’s analysis of the cult of leadership has three elements: (i) companies need to change and adapt; (ii) the CEO, rather than the
shareholders, has power to decide on the direction of change (the strategy);
and, (iii) the CEO may not select the optimal kind of change (for shareholder
value)—it may be overly dramatic. In this paper, we study this idea from the
∗ Dow is at London Business School and Raposo is at ISCTE Business School, Lisbon. This study
was supported by the EU-TMR Research Network Contract FMRX-CT960054. We thank the editor
and the referee for very helpful comments and suggestions. We also thank Aya Chacar, Jan MahrtSmith, Julian Franks, Denis Gromb, and Michael Raith for discussions and written comments
on an earlier draft. We are grateful for comments from seminar audiences at Oxford, Toulouse,
Amsterdam, Pennsylvania, Chicago Graduate School of Business, Northwestern (Kellogg School),
University College London, Pompeu Fabra, Yale School of Management, the European Finance
Association 2001, the American Finance Association, and the Econometric Society Winter Meetings
2002.
2701
2702
The Journal of Finance
perspective of the corporate governance literature, that is, using agency theory.
Note that much of this literature has used an incomplete contracting approach
(Hart (1995)), and each of the three elements in Mintzberg’s argument suggests
contractual incompleteness. This is the approach we take here.
Standard agency theory takes as given the existence of an incentive problem.
The principal and agent agree on a (constrained) optimal contract, and then the
agent goes to work on the problem. One feature of CEO compensation that is
clearly different from this standard agency model is the fact that the contract
is adjusted over time to ref lect the evolution of the firm’s performance and
its strategic direction. There is an annual pay-setting round at which options
and incentive plans are renegotiated. Also, since he or she can inf luence the
firm’s strategy, the CEO may be able to inf luence the compensation contract:
A dramatic merger, or restructuring of the whole business, can lead to larger
options grants.
There are many examples in recent corporate history in which radical corporate change went hand-in-hand with high executive compensation and options
grants: Coca-Cola under Roberto Goizueta, the Daimler–Chrysler merger, GE
under Jack Welch, Chris Gent and the Vodafone-Mannesmann takeover, the
Glaxo Wellcome–SmithKline Beecham merger. Some of these dramatic changes
were successful, others were failures.1 Our paper is not about disastrous dramatic change. It is about change that is expected positive net present value
(NPV) at the outset, given the available information, that may end up being
either successful or unsuccessful, but that is overly dramatic in the sense that
less radical change would have resulted in higher NPV. We argue that the way
executive compensation responds to changes in strategy can lead to management choosing change that is more radical than the shareholders would optimally prefer. For instance, the CEO of a regional electricity utility may find it
personally more profitable to create a global web-based energy market-maker.
We then look at ways in which the incentive contract can be modified at the
outset to anticipate this problem.
This analysis is consistent with (but not identical to) two themes that run
through much of the corporate finance literature, namely, free cash f low theory
and nonvalue-creating mergers. First, there is free cash f low theory. Jensen
(1986, 2000), building on earlier analyses of managerial empire-building (Baumol (1959), Marris (1967), and Williamson (1964)), has argued that managers of
public corporations have a systematic tendency to overinvest.2 Overinvestment
of free cash f low can be viewed as similar to overly dramatic change. For example, cash-generating low-growth businesses may tempt their managers to seek
growth through excessive diversification. Lang and Litzenberger (1989), Lang,
1
Some of the failures can only be criticized with the benefit of hindsight, while others seemed
doomed from the outset.
2
He has even argued that this is so costly that the public corporation is not an efficient institutional vehicle for business ownership, and should be replaced by other institutions (Jensen (1989)).
However, Jensen does not share the perspective offered in this paper that strong incentives may
actually exacerbate managerial conf licts of interest.
CEO Compensation, Change, and Corporate Strategy
2703
Stulz, and Walking (1991), Mann and Sicherman (1991), Blanchard, Lopez-deSilanes, and Shleifer (1994), Kaplan and Zingales (1997), and Lamont (1997)
provide empirical evidence supporting this theory.
Second, there is evidence that many mergers add little or no value to the
acquirer (Asquith, Bruner, and Mullins (1983), Jarrell, Brickley, and Netter
(1988), Bradley, Desai, and Kim (1988), Jarrell and Poulsen (1989), and Bruner
(2002)).3 Mergers are a clear example of dramatic change, and hence, this evidence can help explain why managers may undertake dramatic acquisitions
even when this is not shareholder-value maximizing.
Change is obviously valuable and necessary, and of course we do not argue
that change is intrinsically bad. What we do argue in this paper is that firms
may sometimes change in a suboptimal way. This may be particularly relevant
given that in recent years there has been a high rate of corporate change. The
1990s were a decade of mega-mergers. U.S. merger and acquisition activity
from 1993 to 1999 amounted to an annual average of 8.4% of GDP, compared
with less than 4% in the 1980s and less than 2% in the 1970s (Weston, Siu,
and Johnson (2001), Table 7.4). Moreover, in previous decades merger targets
were typically about 10% of the size of the acquirers, but in the 1990s it became
common for companies to acquire targets almost as large as, or sometimes even
larger than, themselves (AOL-Time Warner and Vodafone-Mannesmann, for
example).
The basic outline of our model is as follows. We assume shareholders are able
to set compensation optimally (subject to incentive compatibility constraints),
but that top management has the advantage of formulating strategy. After the
CEO has chosen a strategy, incentives are set or adjusted before the CEO proceeds to implementation of the chosen strategy. The CEO’s ability to formulate
strategy is part of his job, so this is a natural assumption. We use a setting with
incomplete contractibility4 and limited liability5 for the manager, which leads
to an option-like contract (reward for success, no penalty for failure) conditioned
on firm value.
We assume that change requires substantial effort from the CEO at the implementation stage, while business as usual requires much less. Hence, the reward
for success must induce the CEO to put in enough effort to implement change,
not only in comparison to the alternative of no effort, but also in comparison
to the alternative task of maintaining the status quo. Since implementation of
3
There is some overlap between free cash f low theory and research on nonvalue-creating mergers, since free cash f low may be spent on acquisitions. However, free cash f low may also be spent on
other projects, and many acquisitions are paid for with external finance (new debt or new stock).
Note also that our model can explain mergers that add little value to the acquirer, but cannot
explain mergers that are value destroying.
4
A detailed motivation of the form of contractual incompleteness, and discussion of this hypothesis, is given in the appendix.
5
This feature of CEO compensation is sometimes criticized, but nevertheless it seems to be
almost universal (unlike small private businesses, where the entrepreneur often pledges collateral
to a bank or venture capitalist). In Tirole’s (2001) recent graduate textbook on corporate finance,
all of the models make the same assumption as we do. Studying the economic rationale for this
feature, if there is one, would be interesting but is not the subject of this paper.
2704
The Journal of Finance
a given strategy is noncontractible, we show that the latter constraint is the
binding one. Also, the more personally demanding the alternatives he finds to
the status quo, the higher is his surplus.6 By choosing a task whose success is
highly dependent on his own performance, the CEO is able to extract higher
surplus from his shareholders. We conclude that high-powered incentives can
encourage overly dramatic strategies.7
Anticipating this problem, what could shareholders do to mitigate it? If they
are concerned that the CEO’s incentives are not well aligned with their own,
they could simply give him enough options or equity at the outset and the
conf licts would disappear. In other words, if they are worried that corporate
strategy may be distorted by the CEO so as to extract larger incentive packages of equity or options, they could simply hand over the large package at
the outset. We show that in very unstable environments, characterized by a
high likelihood of change (particularly dramatic change), ex ante contracting might be optimal for shareholders. Thus, high compensation can ref lect
strategic discretion, rather than being proportionate to the CEO’s effort cost of
implementation.
While this approach (“ex ante contracting”) works, there is a tradeoff to be
made: It is expensive. Specifically, the drawback of this type of contract is that
the shareholders commit in advance to high compensation packages as a deterrent to overly dramatic change, even though, ex post: dramatic change might
turn out not to be an option anyway. Assume that the firm’s strategic environment is not completely predictable in advance, and that the CEO has the
advantage of being better informed about it than the shareholders. Then there
is a range of possible strategic choices, not all of which are always available. In
some states of the world, dramatic change is the only positive NPV alternative
to the status quo; in other states, more moderate change is also available (and is
higher NPV than dramatic change), while in yet other states dramatic change is
not an option. In the last case, ex ante contracting means that the CEO ends up
being unnecessarily highly paid to implement a simple strategy. Thus, in environments in which dramatic change is less likely to be an option, shareholders
may prefer a wait-and-see approach to contracting, setting the compensation
package after the strategy is chosen and accepting that this approach might
cause distortion in the strategy-selection process.
Another way for shareholders to discourage excessively dramatic strategies
would be for them to commit in advance not to pay high salaries. It might be
difficult for shareholders to credibly enter into a commitment that ex post they
may want to renegotiate, but perhaps social norms could be the mechanism
that limits CEO pay. We extend the analysis to investigate this case. Such
6
Core and Guay (2002) present empirical evidence that more unstable environments tend to be
positively related to CEO pay–performance sensitivity.
7
High-powered compensation can also cause two other problems: (i) straight cheating by managers; and, (ii) managerial rent-seeking through control of the contracting process. CEOs may use
false accounting to overstate profits and inf late the stock price, and instead of representing shareholders’ interests, boards may cooperate with management in agreeing to excessive compensation.
However, those are not the problems we study here.
CEO Compensation, Change, and Corporate Strategy
2705
commitment works in dissuading the CEO from unnecessary dramatic change,
but again there is a tradeoff as this policy is equally effective in dissuading the
CEO from dramatic change in states in which it is desirable.
How can the conclusions of our model be used to throw light on the evolution
of executive pay in the past 10–15 years? We believe that during this period,
shareholders have become much more aware of the need for large public firms
to change. This is partly due to the revolutionary technological changes that
have taken place. It may also be that shareholders simply became more aware of
this general issue, following a period in the late 1980s during which many firms
that had become insufficiently focused on value creation were restructured by
external means (leveraged buyouts and other hostile takeovers). In terms of
our model, we could hypothesize that before this period, there were informal
political and social conventions that effectively limited CEO pay, and that these
limits became suboptimal because of changes in the parameters of the model,
breaking down as the pressure for change suggested higher and higher CEO
pay. In other words, there was a switch from upper limits on pay, to ex ante
contracting with large incentive pay packages. Another possible explanation
for the changes in compensation practice is that the “difficulty level” of likely
strategies increased over time, which is also consistent with our analysis.
The paper is structured as follows. Section I presents the model and the
main results of the paper. In Section II, we consider extensions in terms of
the contracting mechanism, allowing the firm to set a contract ex ante, or to
precommit to a ceiling in compensation. Section III discusses related empirical
and theoretical literature. Section IV concludes.
I. Modeling Strategy Formulation and Implementation
We assume the firm is run by a risk-neutral manager with no personal financial resources. There are two main steps: The manager formulates a strategy,
and then implements it. (As we will detail in Section A, there is one opportunity
for contracting between these two steps, and there is another opportunity right
at the outset, before the first step.)
Strategy formulation: We assume that the outcome of the strategy formulation process—that is, the strategy that is proposed by the manager—is inf luenced both by the manager’s choice and by random factors. The way we model
strategy formulation is to suppose the CEO is presented with a random menu
of strategies, from which he can pick one only, which he then presents to the
shareholders in addition to the status quo option of no change.
The reason for this assumption is that, because this paper is about incentives
for strategy formulation, we clearly need to assume that the manager’s choice
has a role to play. However, we also need to assume some randomness. If the
outcome depended only on the manager’s choice and there were no uncertainty,
the shareholders would know as much about the firm’s strategic environment
as the manager. In that case, it would be unrealistic to suppose the choice of
2706
The Journal of Finance
strategy would be delegated to the manager.8 It is plausible to assume that the
firm’s strategic environment is better known to the CEO than to the shareholders. Also, developing a strategy takes time and requires a large amount of
attention from the CEO; hence he cannot realistically formulate and present to
the shareholders worked-out plans for implementing all the strategies on the
initial menu—he has to choose.9
To be specific, we assume that with probability p, the menu of strategies observed by the manager contains some alternative to the status quo strategy
B, “business-as-usual.” With probability (1 − p), there is no alternative; we call
this scenario B. If there is an alternative, then the menu can contain a single alternative to B, or two alternatives. This last case presents the manager
with a choice between two strategies as an alternative to B.10 The alternative
strategy may be for moderate change, M, or dramatic change, D. Conditional
on change of some kind being possible, there is a probability qM that moderate
change M will be the only alternative on the menu (we call this scenario BM),
probability qD that dramatic change D will be the only alternative (scenario
BD), and probability 1 − qM − qD that both M and D will be available (scenario
BMD).
When the manager picks one of two alternatives to formulate and present to
shareholders as his strategy, the discarded alternative ceases to be available.
Of course, when the manager presents his strategy, the shareholders do not
know whether it was the only option from the outset, or whether there were
two alternatives. For example, if the manager presents a plan for dramatic
change, the shareholders cannot tell whether this was the only option (scenario
BD) or whether he earlier discarded the option of moderate change (scenario
BMD). If his incentives cause him to prefer dramatic change, the shareholders
can reason that either scenario may have occurred. They can use Bayes’ Rule to
infer the likelihood of each of these two possibilities, and this will be an input
into their choice of optimal contract design—for instance if they themselves
would prefer moderate change.
Strategy implementation: We describe the strategies in terms of parameters,
and impose some conditions on these parameters. Strategy i (for i = B, M, or
D) requires effort ei from the CEO and has a probability π i of success, in which
case the firm is worth V. Otherwise (no effort, or sufficient effort but bad luck),
the firm is worth nothing.
Strategy M, moderate change, involves more effort than carrying out business
as usual B and also has a higher success rate. Likewise, strategy D, dramatic
change, represents a more radical restructuring than M and requires higher
8
And, in any case, the tradeoffs would become trivial. Anytime the shareholders wished to stop
the manager from choosing dramatic change, the “ceiling in compensation” that we describe below
would be a costless way of forcing him to obey their preference.
9
On the other hand, there is no need to formulate a plan merely to stick with the status quo, so
this option always remains available.
10
If there is no alternative to B, the manager cannot report any alternative to the shareholders,
while if there is a single alternative, the manager presents it.
CEO Compensation, Change, and Corporate Strategy
2707
managerial effort. We also assume it has a higher success rate than M. So,
eB < eM < eD and πB < πM < πD . We assume there are diminishing returns to
effort in implementing change
πM
eB
< πB
eM
(1)
and
π B + (π D − π B )
eM − eB
< πM .
eD − eB
(2)
The relation between effort and chance of success is graphed in Figure 1.
These conditions, intuitively, say that the graph looks like a concave function.11 The assumption of diminishing returns is key to our analysis. As we
shall see, the idea that really dramatic change is very costly to implement is
important in deriving our results. Further motivation for this assumption is
given in Appendix A.
Parameter restrictions: We impose further assumptions on the parameters.
First, we assume that
V ≥
eB
.
πB
(3)
This simply means that shareholders are willing to compensate the CEO for
his effort in implementing strategy B; that is, the project is positive NPV in a
first–best sense. For the other strategies, we impose
ei − e B
eB
V−
πi − V −
π B ≥ 0,
(4)
πi − π B
πB
for i = M or D. Condition (4) states that the shareholders will be willing to pay
for the CEO to carry out the strategy for change (M or D) if it is presented
as an alternative to the status quo B. This means that these strategies are
positive NPV in a second-best sense that takes into account a further incentive
compatibility condition, in addition to compensating the CEO for effort (as will
become clear below). This condition is stronger than simply assuming that M
and D are positive NPV in a first–best sense; that is,
ei
eB
V−
πi ≥ V −
πB.
(5)
πi
πB
11
Formally, one cannot say that there is a concave function linking effort and success rate because
the function is not defined on a compact set. It is only defined at effort levels 0, eB , eM , and eD . If
we “fill in” with straight lines between the four points (0, 0), (eB , πB ), (eM , πM ), and (eD , πD ), then the
resulting function is concave.
2708
The Journal of Finance
Figure 1. CEO’s preferences in terms of strategies.
In line with our discussion in the introduction, we are particularly interested
in the case in which M is first–best, that is, the extra chance of success of
strategy D (compared to M) is outweighed by the effort cost:
eM
eD
V−
πM ≥ V −
πD .
(6)
πM
πD
However, this condition is not required as an assumption for the analysis (see
the discussion of Proposition 2 in Section I.B).
A. Contracting between the Shareholders and the Manager
There are three main elements to contracting in our model: The manager’s
limited liability, the assumptions on contractibility, and the timing.
Since the manager has no financial resources, the contract he agrees with
the shareholders can stipulate nonnegative payments only. We assume that
strategies are observable, but not verifiable. Hence, a contract can stipulate
a payment that is conditional on firm value, but not on strategy. We consider
contracts that take the form of a positive payment if the firm is worth V, and
no payment if the firm is worth nothing.
Two key implications of the contractibility assumption are that (i): it is not
possible to make payments conditional on the strategic plans that the CEO proposes to the shareholders and, (ii); the CEO can agree to implement change and
later decide that he prefers to stick to the (easier) status quo. In Appendix A, we
give a detailed discussion and motivation of the noncontractibility assumption.
CEO Compensation, Change, and Corporate Strategy
2709
Potentially, there are two times when contractual arrangements could be
made or renegotiated: After the manager has formulated a strategy but before
he has implemented it, and at the initial stage before he has had time to formulate a strategy. We consider three types of contracts, and study conditions for
each to be optimal. The first type is what we call “ex post contracting.” In this
case, the contract is not set until after the CEO has formulated the strategy.
This allows the contractual terms to be set in response to the strategy that the
manager has formulated. The second type is what we call “ex ante contracting,” where the contract is set at the beginning, but can then be renegotiated
afterwards by mutual agreement, in response to the strategy formulated by
the manager. We then extend the analysis to consider a third type of contracting in which, as part of an ex ante contract, shareholders commit to rule out
renegotiation. Both in the initial negotiation and in the event of renegotiation,
we maximize the principal’s payoff subject to meeting the agent’s reservation
utility level.12
B. Setting CEO Compensation Once Strategy Is Decided: Ex post Contracting
To start, suppose that CEO compensation is not fixed initially. Instead, the
shareholders wait until the CEO has presented his strategy, then they set the
compensation package. In the simplest case (occurring with probability 1 − p),
there is no option for change and the manager’s task is simply to implement
strategy B. Let mB be the contractual payment in the event that the firm is
worth V. Incentive compatibility requires
m B π B − e B ≥ 0,
(7)
and the solution is13
mB =
eB
.
πB
(8)
The CEO obtains his reservation utility level (zero) in this case.
Alternatively, the manager may report that there is an opportunity for change
(either M or D). The resulting payment, conditional on the firm being worth
V, is denoted mi (i = M or D). If the shareholders agree to implement i, then
there are two incentive compatibility conditions. The first requires expected
compensation to outweigh the effort cost,
mi πi − ei ≥ 0,
(9)
and the second requires that the manager really has an incentive to choose i over
B. Since the choice of strategy is not contractible, if the payment is inadequate,
the CEO may pretend to implement i but actually stick to B. The condition is
12
13
This is a standard assumption (Salanie (1997)).
So long as V ≥ mB , as we have assumed in (3). In other words, B is a positive NPV project.
2710
The Journal of Finance
mi πi − ei ≥ mi π B − e B .
(10)
It follows immediately from assumption (1) and (2) that (10) is the binding
constraint; thus the optimal contract overcompensates the manager relative to
his reservation wage (zero). The solution is14
mi =
ei − e B
.
πi − π B
(11)
We denote the CEO’s expected payoff from strategy i as Ui ,
Ui = mi πi − ei
ei − e B
= πB
− eB,
πi − π B
(12)
(13)
where the second equation follows by setting equality in (10) and substituting
(11). We can use this derivation to see the characteristics a CEO likes in a
strategy.
Perversely, the manager prefers strategies with a lower chance of success and
a higher effort level:
PROPOSITION 1: So long as the parameters of the model satisfy condition (4), the
CEO’s preferences for strategies are increasing in effort ei and decreasing in the
chance of success π i .
Proof: Immediate by inspection of (13).
Q.E.D
Furthermore, we can conclude that if the CEO has a choice, he prefers to
formulate strategy D over M : UD > UM , since diminishing returns (2) implies
eD − eB
eB
> πe MM −
. This result can be visualized from Figure 1—given diminishing
πD − πB
− πB
returns to effort, the CEO will always prefer to formulate the more demanding,
difficult-to-implement strategy.
Turning to the shareholders, they prefer both M and D to B, even taking
into account the agency costs of implementing them. Both types of change have
higher NPVs than B (which is itself positive NPV), by assumption (4). However,
shareholders may not agree with the CEO’s unambiguous preference for D
over M. It could be that in scenario BMD, when the manager initially chooses
between M and D, he picks D while the shareholders would prefer M.
The following result characterizes the equilibrium of the model and the conditions under which there is a conf lict of interest between the shareholders
and the CEO. In such a case the CEO chooses to formulate the more complex
strategy D, even if shareholders would have benefitted more from strategy M.
First define
πB
eD − eM
≡ eD − eB − eM − eB
V
(14)
+
πD − πB
πM − πB πD − πM
πD − πM
14
We have assumed in (4) that shareholders are willing to pay that much. In other words, i has
a higher (positive) NPV than B, even allowing for the extra agency costs of implementing it as an
alternative to B.
CEO Compensation, Change, and Corporate Strategy
=
(m D − m M )π B + (e D − e M )
.
πD − πM
2711
(15)
We can now show:
PROPOSITION 2: Suppose that the parameters of the model satisfy conditions (2),
(3), and (4) and the shareholders use ex post contracting.
(i) The CEO presents the hardest strategy i (i = B, M, D) possible, and gets
compensation mi .
(ii) Shareholders agree with this choice except for scenario BMD with V < V̂ .
In that case, they prefer moderate change M while the CEO picks D.
Proof: Part (i): This part of the proposition is immediate because the strategy
choices are trivial, and the compensation contracts have been derived above.
Part (ii): Conditions (2), (3), and (4) ensure that the CEO prefers to formulate
strategy D : UD > UM , as in Proposition 1.
Turning to the shareholders’ preferences, given a straight choice between M
and D, they would prefer M so long as
(V − m M )π M > (V − m D )π D ;
(16)
V (π D − π M ) < (e D − e M ) + (U D − U M ),
(17)
that is,
which, with some algebra, corresponds to V < V̂ . Q.E.D.
Condition (17) has a natural interpretation: The left-hand side is the extra
value created by strategy D; the first term on the right is the extra cost (in terms
of managerial effort) of implementing D, and the second term represents the
additional rents earned by the manager from ex post contracting to implement
D. We conclude that ex post contracting creates a further layer of agency conf lict
in addition to the standard agency problem that the agent is reluctant to exert
effort. The CEO’s strategy choice is distorted because he has an incentive to
seek hard strategies with an unfavorable effort/success ratio, which drive up
the required compensation. He will tend to promote dramatic change D, even
if it is not in the shareholders’ best interests.
Our main concern in this paper, in line with our discussion in the introduction, is the case in which D is “overly dramatic” in the sense that M is first–best
(condition (6)). Because V < V̂ is implied by (6), in this case, there will be a
conf lict of interest between shareholders and the manager using ex post contracting. Define V to be the firm value such that (6) holds with equality; one
can show that V < V̂ . Our main interest is the case V < V . If V becomes
large enough, there is a range (V ≤ V < V̂ ) in which there is still a conf lict of
interest between shareholders and manager, but D is actually first–best. The
divergence between shareholders’ interests and the first–best arises because
2712
The Journal of Finance
they have to pay the manager more than the disutility of effort. If V is larger
still (V̂ ≤ V ), then both parties prefer D.
II. Setting Compensation to Inf luence Strategy Formulation
We now focus attention on the interesting case in which the shareholders and
the CEO disagree over strategies (i.e., V < V̂ as defined in the previous proposition). Anticipating the conf lict of interest with ex post contracting, shareholders
may seek to improve the alignment between their interests and the CEO’s by
designing a better compensation package at an earlier stage. In this section, we
explore two ways for shareholders to correct the CEO’s incentives for strategy
choice, and we discuss the conditions under which each is optimal.
A. Ex Ante Contracting
Clearly, one way of ensuring better alignment of interests is just to give the
CEO a large enough equity share or bonus at the outset: If he gets 99% of firm
value, this difficulty will probably be eliminated. But giving the CEO a large
equity grant, share options, or bonus payment is very expensive. There is a
tradeoff.
We assume that shareholders are able to set the compensation contract ex
ante, before the CEO has started to formulate strategy. Thus, the contract consists of a specified payment in the event that the firm is worth V, and it can be
renegotiated later when the strategy is decided, if doing so is mutually agreeable.15
Suppose that the initial contract is set at m∗ (conditional on the event that
the firm is worth V) and that renegotiation results in final compensation levels
m∗B , m∗M , and m∗D (in the event the firm is worth V) in each of the three possible combinations of feasible strategies that the manager may present to the
shareholders following the strategy formulation stage.16 Define
Ui∗ ≡ mi∗ πi − ei .
(18)
Clearly, m∗B , m∗M , and m∗D cannot be less than m∗ ; otherwise, the manager
would not agree to renegotiate. Also, renegotiation will not take compensation
to a higher level than it would have reached under ex post contracting:
15
We assume here that agents cannot precommit not to renegotiate, an assumption that is
relaxed in Section II.B below.
16
Although the set of possible histories is richer, it is clear that the contract could not distinguish
between events such as: (i) the manager initially had a choice between M and D, but chose to
investigate D; versus, (ii) the manager initially had no choice, and D was the only option for change.
Thus, the most that one could hope for is three final different payments m∗B , m∗M , and m∗D . Even
then, Proposition 3 shows that not all combinations of three positive real numbers are achievable.
CEO Compensation, Change, and Corporate Strategy
2713
PROPOSITION 3: Ex ante contracting results in a floor m∗ , such that
m∗B = max m∗ , m B ,
m∗M = max m∗ , m M ,
m∗D = max m∗ , m D .
Proof: See Appendix B.
Note that ex post contracting can be viewed as a special case of ex ante
contracting: If a low payment is set initially (at most m∗ = mB ), it will simply
be renegotiated upward to m∗B = mB , m∗M = mM , and m∗D = mD , where mB , mM ,
and mD are the ex post contract payments derived in the previous section.
Hence, we refer to this case simply as ex post contracting. We will reserve the
term “ex ante contracting” for the case m∗ > mB .
The idea behind ex ante contracting is as follows (for ease of exposition, the
full derivation is relegated to Appendix B). Suppose the CEO has a choice between strategy D, with compensation mD , and strategy M, with compensation
mM . We know (from Section I.B) that he will prefer D, even though the effort
is higher, because the compensation is much higher. Now increase the compensation for M up to the point at which he is just indifferent: That is, set m∗ so
that
∗
UM
= m∗ π M − e M = m D π D − e D = U D∗ .
(19)
Note that m∗ is lower than mD , because D requires higher effort than M. Also,
m is higher than mM . We give the formula for m∗ in Appendix B. Now, if the
CEO starts off with an ex ante contract m∗ , and it later turns out that there
is no opportunity for change (scenario B), he simply implements B. Likewise,
if the only opportunity for change is M (scenario BM), he will implement M.
If the only opportunity for change is D (scenario BD), renegotiation occurs.
The shareholders will agree to pay mD and the CEO will choose to implement
D. So far, these strategy choices are of course exactly the same as with ex post
contracting; the only difference is that the shareholders have paid more to have
strategies B and M implemented. The interesting case is scenario BMD: Now
the CEO picks M with compensation m∗ , as he has nothing to gain by picking
D instead and renegotiating higher compensation mD .
To determine whether shareholders will prefer ex ante contracting, we can
compute their payoff, conditional on each of the four scenarios, and we can then
compute the expected payoff averaging across the four scenarios and characterize the shareholders’ preference in terms of the exogenous parameters of the
model. This is done in Appendix B. However, more intuitively, we can refer to
the above discussion of the four scenarios. This information is summarized in
the third row of entries in Table I (entitled “ex ante contract”). These entries
show the strategy chosen and the shareholders’ payoff in each scenario. For
comparison, the previous row of the table gives the corresponding information
∗
2714
The Journal of Finance
Table I
Shareholders’ Preferences Toward Different Types of Contracting
in the Four Possible Scenarios of Strategy Availability
The columns correspond to different scenarios of available strategies. The first row (following the
headings in the top row) shows the first–best strategy choice for each of the four scenarios of
available strategy at the strategy formulation stage (this assumes condition (6) from the paper;
otherwise the last column should show strategy D as first–best). The next three rows show the
strategy choice and, underneath it, the shareholder payoff in each of the three contracting environments. Plus and minus signs indicate improvement or worsening relative to previous rows. Note
that mB < mM < m∗ < mD .
Scenario
Available Strategies
B
BM
BD
BMD
First–best strategy
Strategy B
Strategy M
Strategy D
Strategy M
Ex post contract
Strategy B
πB (V − mB )
Strategy M
πM (V − mM )
Strategy D
πD (V − mD )
Strategy D
πD (V − mD )
Ex ante contract:
guaranteed minimum m∗
Strategy B
πB (V − m∗ )
(–)
Strategy M
πM (V − m∗ )
(–)
Strategy D
πD (V − mD )
Strategy D
πM (V − m∗ )
(+)
Precommitment:
upper limit mD
Strategy B
πB (V − mB )
Strategy M
πM (V − mM )
Strategy D
πB (V − mB )
(–)
Strategy D
πM (V − mM )
(++)
for ex post contracting. We can see that compared to ex post contracting, ex ante
contracting makes no difference to the strategy choice in scenarios B and BM,
but results in higher compensation. If these scenarios occur, ex ante contracting is worse for the shareholders. In scenario BD, ex ante contracting makes no
difference to the shareholders, either in the strategy chosen or in pay. Finally,
as shown in the last column, in scenario BMD, ex ante contracting helps the
shareholders: The CEO chooses M instead of D.
This discussion shows that ex ante contracting, that is, commitment to high
payment, will tend to be preferred to ex post contracting if scenarios BMD and
BD are very likely (compared to scenarios B and BM). One could interpret this
as a more changeable environment, or an environment in which a wider range
of challenging strategic options are available. In terms of the parameters of the
model, this corresponds to high p and low qM . The exact conditions are given
in Appendix B.
To summarize, in a highly changeable environment in which difficult strategies are likely, shareholders may guarantee the CEO an apparently overgenerous package of incentives at the outset. Although sometimes overgenerous
relative to the immediate task at hand, it would improve incentives for strategic decision making.
B. Setting a Ceiling on Compensation
In this section, we consider the case in which the shareholders are able
to make precommitments about the ex post renegotiation of the CEO’s
CEO Compensation, Change, and Corporate Strategy
2715
compensation contract by setting an upper limit to compensation. The benefit
of such a compensation scheme is to prevent the CEO from choosing to formulate his preferred strategy D over the shareholders’ preferred strategy M. If he
knows he will never be paid enough to implement an overambitious strategy,
the CEO will lose interest in such a plan. Establishing ex ante a credible ceiling
on compensation does have a drawback, however: Even when dramatic change
is the only possible alternative to the status quo, it will be disregarded.
The issue of precommitment not to renegotiate is open to debate. Clearly, companies and CEOs often do renegotiate compensation. The typical arrangement
consists of an annual salary combined with a stock option grant; so each year
the salary is renegotiated and the CEO also adds more options to an existing
portfolio. Hence, precommitment may seem unrealistic. Also, from a theoretical
point of view, it is hard to see how commitment could be enforced—perhaps by
a third party, but then, the design of the contract with this third party would
have to be complex and might violate the spirit of our prior assumptions on
contractibility between CEO and shareholders, and it would certainly be unrealistic.
However, perhaps social norms or individual ethics could be used as a precommitment device. For example, in Sweden it would probably be possible for
a company to precommit never to pay the CEO over $1 billion (as was paid to
Roberto Goizueta of Coca-Cola over a 10-year period). Even within the United
States, 25 years ago such a payment might have been impossible. Thus, precommitment may be possible at the social level. At the level of the individual
company, it may be possible to precommit by appointing individuals (such as
Warren Buffett) who are known to be strongly opposed to high compensation to
the board or to the remuneration committee. Having large block shareholders
with this view might also have the same effect. For that reason, we believe it is
worthwhile to explore the case in which precommitment is possible. We therefore add the following assumption: We consider the possibility of precommitting
not to renegotiate. The precommitment we consider is an upper bound on the
level of compensation.17
The optimal contract in this case is for shareholders to commit to a maximum
payment mM —in other words, they precommit to never pay enough to implement dramatic change, D (the derivation of the optimal contract is in Appendix
C). The result is that the shareholders agree to payments of mB for strategy B
and mM for strategy M (just as with ex post contracting). In case the CEO proposes strategy D, the shareholders still cannot offer above mB , hence in scenario
BD they agree instead to stick to strategy B with payment mB .
While the full conditions under which this form of contracting is preferred
are given in Appendix C, we now give an intuitive exposition of the tradeoffs
by taking a case-by-case look at each of the four scenarios. To recall, these
scenarios are: (i) Scenario B: strategy B is the only one available; (ii) Scenario
17
This statement rules out more complicated kinds of precommitment, which we consider unreasonable, such as: the shareholders commit never to offer a payment in a set A, and the set A
can have an arbitrary shape such as a subset of the real line that is not connected, the set of rational numbers, etc. Recall that strategies are not contractible, hence the contract cannot be made
contingent on the CEO’s announced strategy.
2716
The Journal of Finance
BM: strategies B and M are available; (iii) Scenario BD: strategies B and D
are available; and, (iv) Scenario BMD: all three strategies are available. The
reader should refer to the last row of Table I, entitled “Precommitment: upper
limit mM ,” and compare it to the previous two rows.
The table shows that setting a maximum wage works well in scenarios B,
BM, and BMD. The optimal precommitment is to a maximum of mM , which
will prevent payment of mD in scenario BMD. More specifically, in scenarios
B and BM, the first–best strategy is chosen, just as with ex post and ex ante
contracting. Moreover, in those two scenarios, the lowest possible compensation is offered, just as it is with ex post contracting but in contrast to ex ante
contracting where there is overpayment. In scenario BMD, again there is an
advantage of setting a limit on compensation: The best strategy is chosen, as it
is with ex ante compensation but unlike ex post compensation, and there is the
further advantage compared to ex ante compensation that this is achieved with
the lowest possible compensation. Finally, in scenario BD, we see the drawback
of setting a limit on compensation: Because strategy D is effectively ruled out,
the firm must revert to the inferior strategy B.
We can see that setting an upper bound would be optimal if it is unlikely that
the strategy for dramatic change, D, is the only one available (low probability
qD ) and if the overall probability of change (p) is also low. The exact conditions
are given in Appendix C. We can interpret this case (low qD and small p) as a
relatively stable environment characterized by a low chance of dramatic change.
In this case, the ceiling on compensation is a cheap way to correct the manager’s
preference for dramatic change. If this is not the case, then it will be preferable
to set either an ex ante contract without a ceiling, or an ex post contract.
III. Related Literature
Our analysis can be used to interpret the available empirical evidence on top
management pay in the United States. Hall and Liebman (1998) document that
on average a 1% increase in a firm’s stock price led to an increase of $124,000 in
the CEO’s wealth in 1994, while in 1998 this value exceeded $500,000. Murphy
(1999) shows a sharp increase in pay–performance sensitivity over the early
1990s, reaching a level almost double the 0.325% reported in the seminal Jensen
and Murphy (1990) article. Conyon and Murphy (2000) report similar findings
for the United Kingdom, showing that both the level and the pay–performance
sensitivity of CEO compensation increased sharply over the 1990s. One can
interpret these studies on the 1990s as evidence of a period of high changeability
and a commitment to high compensation (“ex ante contracting” in terms of our
model).
Core and Guay (2001) describe the cross-sectional variation of executive
compensation, finding that the median large firm has options outstanding
that amount to 5.5% of common stock, rising to 10–14% for growth industries (computer, software, pharmaceutical) but only 2–3% for low-growth industries such as utilities and petroleum firms. Likewise, Murphy (1999) finds
widespread use of stock options in most industry groupings, but not in utilities.
CEO Compensation, Change, and Corporate Strategy
2717
One can interpret this, again, as evidence of higher strategic discretion of
CEOs and more changeable environments in growth industries. An alternative
explanation, in line with standard agency theory, is that higher compensation
in high-growth industries simply ref lects higher agency costs of effort (or private benefits of control). Plausibly, however, while managerial jobs in growth
industries may be more challenging and more onerous, this difference in effort
seems unlikely to be large enough to explain very large divergences in compensation packages. Along the same lines as the findings of the above studies,
Demsetz and Lehn (1985) and Core and Guay (2002) find a strong positive
relation between firm risk and CEO pay–performance sensitivity.18
Core and Larcker (2003) examine firms with “target ownership plans” specifying the minimum amount of stock that must be held by executives and show that
non-CEO executives typically hold much less equity (relative to base salary)
than the CEO. We can interpret this as evidence that given the special role of
the CEO, his incentives should be stronger to give good incentives in strategy
formulation.
We now comment on the relation between this paper and some of the existing
agency literature. First, there is the multitasking version of the principal-agent
problem in which the agent has two tasks to which he can devote effort, but
the output of only one of those tasks is measurable (Holmstrom and Milgrom,
1991). Making incentives more high-powered relative to that output measure
can be counterproductive, reducing the agent’s effort on the other task. There
is a similarity with our model in that incentives for one variable (effort, in our
model) can distort another variable (strategy choice). However, in our model
both variables contribute to the same measured output (firm value). In our
model, the analysis is driven by the incomplete nature of the contracting, the
sequencing of the agents’ choices, and the opportunities for renegotiation of the
contract, all of which are absent from the multitasking model.
Free cash f low theory (Jensen (1986)) suggests that challenging strategies
can be beneficial for CEOs in terms of private benefits. The private benefits are
not explicitly modeled; they are exogenous. Rather than direct private benefits,
we consider indirect benefits arising from the effect of managerial activity on
compensation. Hence, while free cash f low theory is rather broader than our
analysis, our paper could be viewed as compatible with free cash f low theory
and as offering a rationalization of how private benefits can arise.
Shleifer and Vishny (1989) argue that managers have an incentive to entrench themselves by making investments that create specific human capital.
They will favor projects that they alone can operate and that cannot easily be
transferred to another manager, much like a computer programmer who writes
a deliberately cryptic code that makes him or her indispensable and thereby
able to extract rents. In our article, there is no specific human capital. Shleifer
and Vishny’s effects rely on the manager manipulating his position in the labor
market, while the effects in our paper rely on manipulating the agency problem.
18
Core, Guay, and Larcer (2003) contains a detailed discussion of empirical evidence on this
issue and its possible interpretations.
2718
The Journal of Finance
Prendergast (2002) considers a setting in which the agent has discretion in
deciding how to solve the problem. He starts by noting that the majority of
empirical studies find a positive relation between the risk of an agency problem and the strength of the agent’s incentives, contrary to the predictions of
the standard model. This empirical evidence is compatible with our analysis.
He then points out that in reality, agents can decide how to go about solving
a problem, which is similar to our point that CEOs decide strategy as well as
implement it. He goes on to address issues different from the ones we address
here, concluding that if a problem is uncertain, the principal will not know how
to solve it and thus will allow the agent to determine the solution, motivating
him or her with strong incentive pay. On the other hand, if a problem is predictable the principal will know how to solve it, and will instruct the agent with
regard to what to do and then will verify compliance by directly monitoring his
or her actions.
IV. Conclusion
Our results are simple to summarize. In this paper, we analyzed how managerial rent-seeking may distort strategy choice in favor of overambitious change.
We showed that in our model:
r In an environment in which available strategies are likely to include diffi-
r
r
r
cult strategies (leading to disagreement with the CEO), the shareholders
should commit to a policy of high pay, offering the CEO a high-powered incentive package at the outset in order to improve his incentives for strategy
making, even if sometimes this results in excessive (in hindsight) payment
at the implementation stage. This will correct the CEO’s incentives away
from excessively dramatic and hard-to-implement restructuring, and toward the kind of change that is preferred by shareholders.
Another way of curbing a CEO’s tendency toward dramatic change is to
precommit never to pay the very high compensation package required to
implement dramatic change. It may be difficult to make this precommitment, but if it is possible, then it will be preferred if dramatic change is not
likely relative to moderate change, or if change is unlikely overall.
Finally, in some situations it may optimal for the shareholders simply to
accept some strategic distortion and to negotiate compensation to the required level after the strategy has been selected. This will occur when
the existence of alternatives to the status quo is not very likely, or when
shareholders do not have strong preferences regarding the possible change
options.
Our analysis can help to explain why CEOs typically receive much higher
pay and stronger incentives than other senior managers. Their pay ref lects
strategic discretion more than direct compensation for effort. Our results
are also consistent with the empirical evidence of higher incentives in more
changeable environments, and in more recent years. Another possible explanation would be the “difficulty level” of likely strategies increasing over
time.
CEO Compensation, Change, and Corporate Strategy
2719
r One can also hypothesize that the economic pressures in recent years for
firms to transform and reinvent themselves have led to “ex ante contracting,” that is a precommitment to high pay and incentives.
An alternative perspective on CEO pay is that high pay is not optimal at all
for shareholders and simply results from rent extraction in collusion with the
board. While we do not deny this possibility, we think it is worthwhile to explore
how high pay can arise in a model with only relatively minor departures from
standard agency theory.
Appendix A: Remarks on Incomplete Contractibility
The approach we take here applies what is now a standard paradigm, following many papers of the past decade that have studied financial contracting
where some variables are observable by both parties, but are not contractible.
Aghion and Bolton (1991), Hart (1995), and Hart and Moore (1990) are leading
examples. In general, noncontractibility can arise because it would be too expensive or too complex to make contracts fully conditional, or because it would
be difficult for a third party to verify fulfillment of the conditional clauses in the
contract (since contracts depend on third parties such as courts or private arbitrators for enforcement). For example, Hart (1995) and Hart and Moore (1998)
assume that investment is not contractible. They argue that even if third parties were able to verify monetary expenditures on investment, they would be
unable to tell if the funds were applied properly to the right kinds of projects.
This is similar to our assumption that strategy is noncontractible. Only “hard”
variables are contractible.
To motivate our analysis, consider the following example. Suppose that a professor is given responsibility for a degree program. He or she can choose to run
the program on a business-as-usual basis, that is, attend the usual committee
meetings, check the lecturing performance of his or her colleagues, monitor the
performance of the program office in regard to admissions and records, and take
an appropriate interest in student welfare. This is a moderately time consuming, but manageable task and the professor would be entitled to expect some
measure of compensation in the form of a reduced teaching load or a moderate
salary supplement.
Now suppose the professor is tasked with restructuring the course with a
freshly thought-out structure and syllabus that ref lects current student demand and that is fully up to date. The first observation we make is that
engineering such a change is extremely demanding in terms of time and energy. Many proposed changes, however trivial, will meet with resistance from
entrenched lecturers. Even colleagues who are enthusiastic about change will
respond by making a wide variety of counterproposals that make coordination
on an improved outcome difficult. So, change from the status quo can be disproportionately costly and the professor, unless he obtains intrinsic satisfaction
from the task, will not be keen to take on this job without substantial additional
compensation (generally in universities, such extra compensation is unlikely!).
2720
The Journal of Finance
Of course, firms may not be quite as resistant to change as some universities
but we argue that the same broad feature applies.
The second observation we make is that it may be easy to spot whether the
changes made by a colleague are profound or superficial, but it may be hard to
prove that assessment in a way that is credible to an outsider. A professor may
feel that the course director’s “relaunched” program is just a minor variant
of the old one, but, if challenged on this view, it might be hard for him or
her to convince a colleague in another department, a student, or the course
director’s lawyer. So it would be impossible to condition compensation on the
successful implementation of a truly improved degree program. We argue that
in businesses generally, similar reasons are likely to make it impossible to
condition pay directly on strategy. However, in businesses, unlike universities,
pay can be conditioned on share price.
To summarize the two key points of the example: First, change can require
disproportionately high effort to implement, relative to the status quo; second,
writing a contract conditional on “satisfactory change” is likely to be impossible.
In our model, we have in mind a variety of corporate strategies, of which a possible simple example would be the introduction of a comprehensive cost-cutting
program. At first sight, it might seem that costs can be verified easily from
the company accounts. However, this would be a crude way of monitoring the
successful implementation of a genuinely shareholder-value increasing rationalization plan. It could likely be quite easy for the manager to slash costs with
a poorly executed program of cutbacks that would actually damage shareholder
value in the longer term (through employee morale, quality control problems,
disruption of supplier relationships, etc). Just because a strategy has implications for an easily quantifiable variable, it does not mean that controlling that
variable will actually verify implementation of the strategy.
Holmström and Milgrom (1991) give a good illustration of this kind of problem. A teacher in South Carolina was found to have boosted her class exam
performance, and hence her own performance rating, by passing the children
answers to the statewide geography test. Another egregious example is given
in the Financial Times of February 4, 1994: The recently appointed chairman of
Audi discovered that the previous year’s sales figures had been “bolstered by an
old trick. Audi France officials confirmed yesterday that ‘several tens of thousands’ of cars had been parked with French distributors [and hence recorded
as sales], only to be shipped back to Germany last year. Since many of them
lacked airbags and ABS (antilock brake system) braking systems—regarded
as essentials in the Germany quality car market—selling them was no easy
task.” This example of “channel stuffing” shows why even apparently “hard”
data such as company accounts may not be of much help in writing conditional
contracts.
Appendix B: Ex Ante Contracting
This section of the appendix contains full details and derivations for the
analysis given in Section II.A of the main text. The correct standard procedure
CEO Compensation, Change, and Corporate Strategy
2721
for solving an agency problem requires us to list all the possible actions of the
agent, for each action to compute the cheapest way for the principal to induce
the action, and then to compute the principal’s resulting payoff. Finally, the
principal must compare his payoff across actions and pick the optimal one. For
this model, the action space of the agent is complex because the CEO needs to
decide whether to choose M or D, if there is a choice, at the strategy formulation
stage; whether to choose B or the alternative, if there is one, at the strategy
choice stage; and whether to put in the required effort for the chosen strategy
at the implementation stage. However, we can simplify matters by eliminating
actions such that the CEO does not put in effort (as shown in Appendix D),
and can specify the relevant actions by describing the choices of strategy at the
formulation stage and at the implementation stage. The list and full description
of the eight such actions is given in Appendix D, where we also show that six of
these can be eliminated from consideration here (ex ante contracting) because
they involve ex post inefficient strategy choices. The two remaining actions are
what we call:
r Moderate Change Favored: The manager selects M over D if there is a choice
r
at the strategy formulation stage (scenario BMD), and this is then chosen
for implementation over B. If M is the only alternative at the strategy
formulation stage (scenario BM), it is chosen for implementation over B,
likewise for D (scenario BD).
Dramatic Change Favored: The manager selects D over M in scenario BMD
at the strategy formulation stage, and this is then chosen for implementation over B. In scenario BM, M is chosen for implementation over B,
likewise for D in scenario BD.
PROPOSITION 4: The cheapest way to implement Dramatic Change Favored is
ex post contracting.
Proof: Ex post contracting gives this action in return for contractual payments (in the event of firm value V) m∗B = mB , m∗M = mM , m∗D = mD . If any of
these payments were reduced, the manager would not be willing to work to
implement the specified strategy B, M, or D; therefore, this is the cheapest way
to induce the action. Q.E.D.
The next proposition will rely on Proposition 3 from the main text, so we first
give that proof.
Proof of Proposition 3: We show that m∗B = max{m∗ , mB }; a similar argument
holds for the other two cases, BM and BD. Let the ex ante contract specify a
payment m∗ in the event the firm succeeds and is worth V. First note that if
renegotiation occurs, it will occur upward only, otherwise the CEO will reject
the proposal. Hence m∗B ≥ m∗ . Next, note that the renegotiated payment cannot
exceed the ex post contracting level, otherwise the shareholders could offer less
and achieve the same behavior from the CEO.
Suppose first that m∗ ≥ mB , then it is clear we cannot have renegotiation and
∗
mB = m∗ .
2722
The Journal of Finance
Next suppose that m∗ < mB , then it is clear (from the derivation of mB ) that
renegotiation is mutually beneficial and will lead to a contractual payment
m∗B = mB . Q.E.D.
PROPOSITION 5: The cheapest way to induce Moderate Change Favored is ex ante
contracting with m∗ ∈ (mM , mD ) given by
m∗ =
πD
eD − eB
− (e D − e M )
πD − πB
.
πM
(B1)
Proof: To induce this action, we obviously cannot drop m∗B below mB , m∗M
below mM , or m∗D below mD , otherwise the CEO will not be willing to work to
implement the chosen strategy. We have to increase one or more of the contractual payments above the ex post level by setting m∗ > mB , to see if this will
induce the agent to pick M over D and if so, find the smallest m∗ that does this.
The question is whether it is optimal to have m∗ ∈ (mB , mM ], m∗ ∈ (mM , mD ], or
m∗ > mD .
In the event the CEO has a choice between M and D at the strategy formulation stage, his expected payoff if he picks M is (πM m∗M − eM ); if he picks D,
it is (πD m∗D − eD ). Since ex post contracting induces the choice of D, it follows
that m∗ ∈ (mB , mM ] will not induce the choice of M because increasing m∗ from
mB part-way toward mM does not affect either of these payoffs. Next, consider
m∗ > mM . If m∗ ∈ (mM , mD ], the CEO will just be willing to pick M over D if
(B2)
π M m∗M − e M = π D m∗D − e D .
By hypothesis m∗M = m∗ and m∗D = mD , so substituting for m D =
have
eD − eB
− (e D − e M )
πD
πD − πB
m∗ =
.
πM
eD − eB
,
πD − πB
we
(B3)
One can verify that m∗ defined by this formula does not exceed mD , as a consequence of our assumption that eeMD > ππMD (or equivalently, mM < mD ). Q.E.D.
We can now complete the solution of the problem by examining the shareholders’ preferences. If the shareholders induce choice of Moderate Change Favored
using ex ante contracting with m∗ as just derived, their expected payoff (averaging across all four scenarios) is
(1 − p)(V − m∗ )π B + p(1 − q D )(V − m∗ )π M + pq D (V − m D )π D .
(B4)
If instead they just accept ex post contracting to induce choice of Dramatic
Change Favored, their expected payoff is
(1 − p)(V − m B )π B + pqM (V − m M )π M + p(1 − qM )(V − m D )π D.
(B5)
CEO Compensation, Change, and Corporate Strategy
2723
It follows immediately that:
PROPOSITION 6: The shareholders will use ex ante contracting with a guaranteed
minimum m∗ (paid in the event of success) to induce the CEO to prefer Moderate
Change Favored if
p(1 − qM − q D )(V − m∗ )π M ≥ p(1 − qM − q D )(V − m D )π D
+ (1 − p)(m∗ − m B )π B
+ pqM (m∗ − m M )π M .
(B6)
This ex ante contracted payment m∗ will be renegotiated upward to mD in the
event that the CEO has no choice over available strategies and presents D to the
shareholders.
If the inequality is reversed, they will prefer the ex post contract implementing
Dramatic Change Favored.
This expression has a simple intuitive explanation. The term on the left-hand
side represents the shareholders’ payoff when the CEO chooses to formulate
M over D, which is the benefit of a regime of ex ante contracting. For this
to be desirable, it must be greater than the payoff when the CEO makes the
opposite choice (the first term on the right-hand side (RHS)). But it must also
compensate the shareholders for two drawbacks associated with ex ante contracting: (i) the shareholders will have to overpay the CEO (relative to the ex
post levels) for implementing strategy B when there are no alternatives to it
(the second term on the RHS); and, (ii) they will also overpay the CEO (relative
to the ex post levels) for implementing strategy M, even when dramatic change
D is not a real threat (third term on the RHS).
Appendix C: Setting a Ceiling on Compensation
This section of the appendix contains full details and derivations for the
analysis given in Section II.B of the main text. We consider an ex ante contract with a precommitment to an upper limit for compensation; during the
renegotiation stage, this upper limit will be binding. Clearly, this will not help
implementation of the two actions previously analyzed in Appendix B (Moderate Change Favored and Dramatic Change Favored), because they require
a high payment for strategy D. The interesting case now is the possibility to
implement another action which we call Moderate Change Only.19 This action
is characterized by choosing M whenever it is available and ignoring D even
if it is the only alternative to B. The shareholders can correct the manager’s
preference for D by committing themselves never to pay enough compensation
to make the CEO willing to implement that strategy. This therefore requires
mB ≥ mB , and mM ≥ mM but not mD ≥ mD (where mi is the final payment made
19
Denoted by action A7 in Appendix D. In that Appendix, we also show that the remaining five
actions, A2, A3, A4, A6 and A8, are not optimal.
2724
The Journal of Finance
to the CEO in the event of success given that he chooses to implement strategy
i). So the cheapest way to implement this action is to precommit to never pay
more than mM . Hence, mB = mB , mM = mM , and mD = mM . The expected payoff
for the shareholders is
(1 − p)π B (V − m B ) + pq D π B (V − m B ) + p(1 − q D )π M (V − m M ).
(C1)
We can compare this expected payoff to what shareholders would obtain in
the case that the contracting process involved ex ante contracting without precommitment, and also in the case that we had an ex post contracting policy.
From these comparisons we reach the following results.
PROPOSITION 7: The shareholders will prefer ex ante contracting with a ceiling
on compensation (inducing the CEO to choose Moderate Change Only) to ex ante
contracting with unrestricted renegotiation if:
π B (1 − p) m∗ − m B + p(1 − q D )π M m∗ − m M ≥ pq D [π D (V − m D )
− π B (V − m B )],
(C2)
with m∗ as defined in Proposition 5.
Proof : Direct from the comparison of the shareholders’ expected payoff with
ex ante contracting with a ceiling (equation (C1)) and without an upper bound
(equation (B4)). Q.E.D.
PROPOSITION 8: The shareholders would prefer the ex ante contract with a ceiling
(inducing the CEO to choose Moderate Change Only) to ex-post contracting if:
pq D [(V − m B )π B − (V − m D )π D ] ≥ p(1 − qM − q D )[(V − m M )π M
− (V − m D )π D ].
(C3)
Proof: Direct from the comparison of the shareholders’ expected payoff with
ex ante contracting with a ceiling (equation (C1)) to their expected payoff with
ex post contracting (equation (B5)).
If both conditions (C2) and (C3) are satisfied, the shareholders would prefer to
rule out the possibility of the manager formulating strategy D by precommitting
to a ceiling on compensation. Q.E.D.
Appendix D: List of Actions
We start by showing that we only need to consider actions such that the CEO
puts in effort.
LEMMA 1: Under an optimal contract, the CEO always puts in the required effort
for the chosen strategy.
Proof: Consider first the case of ex post contracting and the possibility that
in the event that B is the only available option, the CEO does not put in the
CEO Compensation, Change, and Corporate Strategy
2725
required effort, eB (implying m∗B < mB ). By our assumption that V > mB , this is
suboptimal because the shareholders would prefer to pay mB and induce effort.
The cases BM and BD are similar given our assumptions that the parameters
satisfy condition (4). In the case of ex ante contracting, the same reasoning
applies.
In the case of a ceiling on compensation, we know that the ceiling would
never be below mB , by our assumption that V > mB . Therefore, if the manager
moves away from implementing B with effort, it must be in order to implement
another strategy with effort. Q.E.D.
We can list the actions as follows:
A1: (D, MBM , DBD ) Dramatic Change Favored
A2: (D, BBM , DBD )
A3: (D, MBM , BBD )
A4: (D, BBM , BBD )
A5: (M, MBM , DBD ) Moderate Change Favored
A6: (M, BBM , DBD )
A7: (M, MBM , BBD ) Moderate Change Only
A8: (M, BBM , BBD ).
To explain the notation, take the first action on the list, (D, MBM , DBD ), which
represents the outcome of ex post contracting. The first symbol, D, means that
when the manager has a choice between M and D at the strategy formulation
stage (scenario BMD, with probability p(1 − qM − q D )), he picks D. The second
symbol, MBM , means that at the strategy implementation stage when there is a
choice between B and M, he picks M—note that this is relevant because even if
he picks D in preference to M at the strategy formulation stage when he has the
choice, there will sometimes (scenario BM, with probability qM ) be occasions
when M is the only option. The third symbol, DBD , means that at the strategy
implementation stage when there is a choice between B and D, he picks D.
The three actions that are referred to in the text are A1 (Dramatic Change
Favored), A5 (Moderate Change Favored), and A7 (Moderate Change Only).
For the purposes of the analysis in Sections I.B and II.A (i.e., the main analysis on ex post contracting and ex ante contracting), we can rule out actions A2–
A4 and A6–A8 as suboptimal, because renegotiation would always take place to
implement the change strategy (M or D) rather than the status quo B by our assumptions on the exogenous parameters satisfying condition (4). The argument
is similar to the previous proposition. That leaves action A5 (Moderate Change
Favored, which corresponds to the manager choosing to investigate M rather
than the more dramatic D) in addition to A1 (Dramatic Change Favored).
For the extension to the analysis of precommitment in Section II.B, we cannot
exclude actions simply because they are ex post suboptimal—after all, the purpose of precommitment is precisely to improve ex ante incentives by allowing
ex post inefficient actions in some outcomes. However, one can readily see that
A2 and A6 are inferior to A1, A3 is inferior to A7, and A4 and A8 are inferior
to all three actions A1, A5, and A7. That leaves A1, A5, and A7 as analyzed in
the previous sections.
2726
The Journal of Finance
REFERENCES
Aghion, Philippe, and Patrick Bolton, 1991, An incomplete contracts approach to financial contracting, Review of Economic Studies 59, 473–494.
Asquith, Paul, Robert Bruner, and David Mullins, 1983, The gains to bidding firms from merger,
Journal of Financial Economics 11, 121–139.
Baumol, William J., 1959, Business Behavior, Value and Growth (Macmillan, New York).
Berle, Adolf, and Gardiner Means, 1932, The Modern Corporation and Private Property (Macmillan,
New York).
Blanchard, Olivier Jean, Florencio Lopez-de-Silanes, and Andrei Shleifer, 1994, What do firms do
with cash windfalls? Journal of Financial Economics 36, 337–360.
Bradley, Michael, Anand Desai, and E. Han Kim, 1988, Synergistic gains from corporate acquisitions and their division between the stockholders of target and acquiring firms, Journal of
Financial Economics 21, 3–40.
Bruner, Robert, 2002, Does M&A pay? A review of the evidence for the decision-maker, Journal of
Applied Finance 12, 48–68.
Conyon, Martin, and Kevin Murphy, 2000, The prince and the pauper? CEO pay in the US and
UK, Economic Journal 110, 640–671.
Core, John, and Wayne Guay, 2001, Stock option plans for non-executive employees, Journal of
Financial Economics 61, 253–287.
Core, John, and Wayne Guay, 2002, The other side of the trade-off: The impact of risk on executive
compensation: A revised comment, Working paper, University of Pennsylvania and SSRN.
Core, John, and David Larcker, 2002, Performance consequences of mandatory increases in executive stock ownership, Journal of Financial Economics 64, 317–340.
Core, John, Wayne Guay, and David Larcker, 2003, Executive equity compensation and incentives:
A survey, Economic Policy Review 9, 27–50.
Demsetz, Harold, and Kenneth Lehn, 1985, The structure of corporate ownership: Causes and
consequences, Journal of Political Economy 93, 1155–1177.
Hall, Brian, and, Jeffrey Liebman, 1998, Are CEOs really paid like bureaucrats? Quarterly Journal
of Economics 103, 653–691.
Hart, Oliver, 1995, Firms, Contracts and Financial Structure (Clarendon Press, Oxford).
Hart, Oliver, and John Moore, 1990, Property rights and the nature of the firm, Journal of Political
Economy 98, 1119–1158.
Holmstrom, Bengt, and Paul Milgrom, 1991, Multi-task principal agent analyses: Incentive contracts, asset ownership and job design, Journal of Law, Economics and Organization 7, 24–51.
Jarrell, Greg, James, Brickley, and Jeffry Netter, 1988, The market for corporate control: The
empirical evidence since 1980, Journal of Economic Perspectives 2, 49–68.
Jarrell, Greg, and Annette Poulsen, 1989, Returns to acquiring firms in tender offers: Evidence
from three decades, Financial Management 18, 12–19.
Jensen, Michael C., 1986, Agency costs of free cash f low, corporate finance, and takeovers, American
Economic Review 76, 323–329.
Jensen, Michael C., 1989, The eclipse of the public corporation, Harvard Business Review
(September-October), updated version, 1997, available on SSRN.
Jensen, Michael C., 2000, Theory of the Firm: Governance, Residual Claims, and Organizational
Forms (Harvard University Press, Cambridge, MA).
Jensen, Michael C., and Kevin J. Murphy, 1990, Performance pay and top-management incentives,
Journal of Political Economy 98, 225–264.
Kaplan, Steven, and Luigi Zingales, 1997, Do financing constraints explain why investment is
correlated with cash f low? Quarterly Journal of Economics 112, 169–215.
Lamont, Owen, 1997, Cash f low and investment: Evidence from internal capital markets, Journal
of Finance 51, 83–109.
Lang, Larry, and Robert Litzenberger, 1989, Dividend announcements: Cash f low signalling vs.
free cash f low hypothesis? Journal of Financial Economics 24, 181–192.
Lang, Larry, René Stulz, and Ralph Walking, 1991, A test of the free cash f low hypothesis, Journal
of Financial Economics 29, 315–335.
CEO Compensation, Change, and Corporate Strategy
2727
Mann, Steven, and Neil Sicherman, 1991, The agency costs of free cash f low: Acquisition activity
and equity issues, Journal of Business 64, 213–227.
Marris, Robin, 1967, The Economic Theory of Managerial Capitalism (Macmillan Londan).
Mintzberg, Henry, 2000, View from the top: Henry Mintzberg on strategy and management, The
Academy of Management Executive 14, 31–39.
Murphy, Kevin J., 1999, Executive compensation, in O. A. Ashenfelter and D. Card, eds.: Handbook
of Labor Economics, Vol. 3 (North Holland, Amsterdam).
Prendergast, Canice, 2002, The tenuous trade-off between risk and incentives, Journal of Political
Economy 110, 1071–1102.
Salanié, Bernard, 1997, The Economics of Contracts (MIT Press, Cambridge, MA).
Shleifer, Andrei, and Robert W. Vishny, 1989, Management entrenchment: The case of managerspecific investments, Journal of Financial Economics 25, 123–140.
Shleifer, Andrei, and Robert W. Vishny, 1997, A survey of corporate governance, Journal of Finance
52, 737–783.
Tirole, Jean, 2001, Lecture notes on corporate finance, unpublished manuscript, University of
Toulouse.
Weston, J. Fred, Juan A. Siu, and Brian A. Johnson, 2001, Takeovers, Restructuring and Corporate
Governance, 3rd. edition (Prentice-Hall Upper Saddle River, NJ).
Williamson, Oliver, 1964, The Economics of Discretionary Behavior: Managerial Objectives in a
Theory of the Firm (Prentice-Hall, Englewood Cliffs, NJ).
2728